1. Field of the Invention
The present invention relates to a target position detection apparatus for a robot, and particularly to a target position detection apparatus adapted for teaching the robot about an accurate position.
2. Background Art
In the fields of the semiconductor device and liquid crystal display device, it has become more and more difficult to teach the robot, as the device has been more complicated and the size of each part or article to be carried in a process has been more increased. To teach the robot about an accurate position is very important for reliability of the robot. In such a situation, incorrect or inadequate teaching provided to the robot due to a lack of skill of an operator may lead to a serious problem. Therefore, there is a desire for a technique that can teach the robot about the accurate position, without depending on any special skill of the operator.
Around the robot, peripheral devices, such as stages and/or other mechanical parts, etc., are installed. For the teaching to the robot, it is necessary to know or judge whether or not the robot and its peripheral devices are adequately positioned and/or correctly installed relative to one another. Therefore, a further technique for automatically performing such judgment is also desired.
U.S. Pat. No. 6,242,879 (Patent Document 1) discloses a technique for obtaining a teaching point by detecting a position of a target by a three-axis-SCARA type robot. FIG. 13 is a front view of a conventional robot 1, and FIG. 14 is a plan view of the conventional robot 1. In the technique disclosed in the Patent Document 1, an end effector 2 of the robot 1 is moved toward the target provided to a cassette 3 so as to be in contact with the target. At this time, changes in the torque and speed are detected, respectively. Thus, each contact point between the end effector 2 and the target can be detected, by comparing the changes in the torque and speed in the case in which the end effector 2 is contacted with the target and those in the case in which the end effector 2 is not contacted with the target. In this way, the position of the target can be obtained from each detected contact point, as such the teaching point can be calculated.
In the technique disclosed in the Patent Document 1, deformation of the end effector and/or target and occurrence of particles may be seen when the end effector 2 is contacted with the target. In order to avoid such inconveniences, it is necessary to operate the robot 1 at a considerably low speed. In such a case, however, deterioration of accuracy in positional detection may occur since the fluctuation factors and time-dependent changing factors in a driving system for the robot 1 will become dominant. The fluctuation factors include fluctuation of torque, friction and the like. The time-dependent changing factors include hysteresis and the like.